Professional Documents
Culture Documents
https://doi.org/10.1007/s00228-019-02713-1
Received: 15 May 2019 / Accepted: 25 June 2019 / Published online: 4 July 2019
# Springer-Verlag GmbH Germany, part of Springer Nature 2019
Abstract
Purpose Tendinopathy is a known adverse reaction associated to fluoroquinolones, but a meta-analysis was not yet published.
The aim of this study was to conduct a systematic review and a meta-analysis of the scientific evidence evaluating the risk of
tendon injury associated with fluoroquinolones.
Methods A literature search was conducted to identify observational studies which reported results on the risk of Achilles tendon
rupture (ATR), risk of Achilles tendinitis (AT), or risk of any tendon disorders (ATD). A meta-analysis was performed by pooling
odds ratios (ORs) with their 95% confidence intervals (CIs).
Results Fifteen studies were included in the meta-analysis. Treatment with fluoroquinolones was associated with an increased
risk of ATR (OR 2.52 (95% CI 1.81–3.52), p < 0.001, I2 = 76.7%), an increased risk of AT (OR 3.95 (95% CI 3.11–5.01),
p < 0.001, I2 = 0%), and increased risk of ATD (OR 1.98 (95% CI 1.62–2.43), p < 0.001, I2 = 84.5%). The initial risk estimates
remained statistically significant among patients aged ≥ 60 years old. Risk estimates did not significantly change after depending
on the concomitant use of corticosteroids or studies methodological quality assessment. The analysis according to the type of
fluoroquinolones was only possible for ATR, which were ofloxacin and norfloxacin were found to increase the risk of this
outcome, but not ciprofloxacin and levofloxacin.
Conclusions The results of this meta-analysis confirm the risk of tendon injuries associated with fluoroquinolones. Older age and
concomitant use of corticosteroids seem to be additional risk factors for tendinopathy.
In 2016, the US FDA reinforced warnings about the risk of associated with systemic fluoroquinolones. MeSH and
tendinopathies associated with the use of systemic Emtree terms, the MedDRA dictionary (v.21.0), and the
fluoroquinolones [6]. Since the risk of these serious adverse International Nonproprietary Names (INN) of
effects may outweigh the benefits of systemic fluoroquinolones were used in the search equation [20–22].
fluoroquinolones in the treatment of minor infections, these Only the literature published in the English language was
antibiotics should be reserved for use when no alternative considered for inclusion. Bibliographic references list of all
treatment options can be considered [6]. In 2018, the relevant studies, meta-analyses, and systematic reviews were
European Medicines Agency (EMA) also recommended hand searched in order to identify additional eligible articles.
new restrictions on the use of systemic fluoroquinolones after We did not seek to identify safety data of fluoroquinolones
reviewing safety information about disabling and potentially beyond published studies. The electronic databases search
long-lasting adverse effects, including tendon injuries [8, 9]. strategy is available in Supplemental Table 1.
Some systematic reviews of studies evaluating the
risk of tendinopathies associated with the use of system-
atic fluoroquinolones have been carried out [2, 10–12]. Study selection, data extraction, and quality
However, important limitations seem to characterize assessment
those works. Not all the observational studies available
in the scientific literature were conducted in the previ- The titles and abstracts of all retrieved citations were screened
ous systematic reviews, particularly the most recently by two independent reviewers (C.A. and D.M.) to identify
published evidence [9, 13–15]. Moreover, although a potentially relevant publications. Full texts were retrieved for
meta-analysis has been conducted, the results were not relevant citations. Discrepancies were resolved by majority
stratified according to the different outcomes assessed in decision (two out of three) involving a third investigator
the epidemiological studies, different risk windows, (F.B.M.). Studies would be included if they fulfilled the fol-
studies methodological quality, or type of comparators lowing criteria: 1, being a comparative observational study
used [12]. (case-control or cohort studies); 2, included patients of all ages
Since fluoroquinolones are essential medicines, it is and genders; 3, compared fluoroquinolones with a placebo (or
important, from a public health perspective, to evaluate non-use) or active control; and 4, have provided risk estimates
the consistency of the results from observational studies (relative risk (RR), odds ratio (OR), or hazard ratio (HR)) or
assessing the risk of tendinopathies. Although meta- data allowing to calculate such risk estimates. Only peer-
analyses may have some limitations, quantitative synthesis reviewed full papers were considered. Basic science studies,
is usually preferred over qualitative analysis of the results reviews, case reports, and studies without a comparison group
of a set of studies [16]. Moreover, conducting a meta- were excluded.
analysis may allow assessing the consistency of the results The following data from the studies was extracted: design,
and understanding the influence of risk factors and addi- characteristics of participants, treatments under evaluation,
tional variables in the risk of tendinopathies associated risk-window length, and estimated effect measures.
with the use of fluoroquinolones [17, 18]. Studies’ methodological quality was assessed through
The aim of this systematic review and meta-analysis is to Newcastle–Ottawa scale, which considers the following char-
assess the risk of tendon injuries associated with the use of acteristics: selection of the study groups, comparability of the
systemic fluoroquinolones, as well as to explore risk varia- groups, ascertainment of either the exposure (for case-control
tions due to risk factors, study designs, and methodological studies), or outcome of interest (for cohort studies) [23]. A
quality of included studies. maximum of one point for each item within the “Selection”
and “Exposure/Outcome” categories could be awarded. For
“Comparability” category, a maximum of two points could
Methods be awarded. The summary score equals the number of points
earned by each study, totaling a maximum of 9 points. Studies
The current systematic review and meta-analysis were report- scoring ≥ 8 points were considered to be of good quality, those
ed according to the Preferred Reporting Items for Systematic scoring < 8 and ≥ 6 points were of moderate quality, and those
Reviews and Meta-Analyses (PRISMA) statement [19]. scoring < 6 points were considered to have poor quality. Two
independent reviewers assessed the methodological quality of
Literature search the studies (C.A. and D.M.). A third investigator (F.B.M.)
helped to resolve discrepancies by majority decision. When
PUBMED, EMBASE, and Cochrane Library were searched more than one reference was found for the same study, the
from its inception until November 2018 in order to identify methodological quality evaluation was based on the total set
relevant studies evaluating the risk for tendon injury of information.
Eur J Clin Pharmacol (2019) 75:1431–1443 1433
Outcomes assessed analysis comprising seven cohort studies [15, 27–32] and
eight case-control studies [9, 13, 14, 33–37].
The following outcomes were assessed in this meta-analysis:
risk of Achilles tendon rupture (ATR), risk of Achilles tendi-
Studies characteristics
nitis (AT), and risk of any tendon disorders (ATD). The out-
comes selected for this systematic review and meta-analysis
The main characteristics of the studies are presented in
were the same as those considered in a previous systematic
Table 1. Study design, participants’ demographic characteris-
review [10]. The ATR and AT are the most frequently evalu-
tics, drugs under evaluation, and a number of subjects evalu-
ated tendon injury outcomes in published studies. The ATD is
ated were described. Patients’ mean age varied from 43.5 to
a composite outcome of any tendon rupture and/or any
65.0 years and the proportion of female patients included in
tendinitis.
the studies ranged from 22.1 to 71.4%. Most of the studies
elected the population sample from either prescription dis-
Statistical analysis pensing data, health administrative data, or general practice
data, although one study examined a post-transplant popula-
A meta-analysis was performed by pooling odds ratios (ORs) tion (heart) and one study evaluated patients with end-stage
with their 95% confidence intervals (CIs), using the renal disease [30, 32]. Only one study used propensity score to
DerSimonian and Laird random-effects model [24]. This match participants [31], while the remaining adjusted the anal-
model was chosen since the validity of tests of heterogeneity yses for several variables. Five studies compared the risk of
can be limited with a small number of component studies and tendon injuries in fluoroquinolone users with active controls
it is more conservative than a fixed effect model in the pres- (other antibiotics), while the remaining compared fluoroquin-
ence of between-studies heterogeneity. If the studies presented olone users with non-users [15, 27–29, 31]. Thirteen studies
more than one risk estimate, the most adjusted one would be defined a risk window of 30 days between fluoroquinolone
used. The effect size estimates available for the shortest time exposure and outcome development. Two studies adopted a
intervals between fluoroquinolones prescription and tendon 90-days risk period and the other two did not report the length
injury diagnosis or treatment were used in the overall risk of time during which the outcome was assessed. Nine studies
estimate. The I2 statistic test was used to assess heterogeneity assessed the risk of ATR, three the risk of AT, and eight the
between studies. An I2 estimate > 50% was considered indic- risk of ATD.
ative of substantial heterogeneity [25]. The publication bias The results of the assessment of studies’ methodological
was visually examined by a funnel plot and statistically eval- quality are also presented in Table 1. The methodological
uated by Egger’s regression asymmetry test [26]. quality was assessed as “good” for three studies, “moderate”
A sensitivity analysis was conducted to explore the influ- for nine, and “poor” for three. Lack of information on con-
ence of the following variables on the summary estimates: founding variables used to adjust risk estimates was the main
study designs, studies’ methodological quality scores, risk methodological impairment found among these studies.
windows defined in included studies, age (≥ 60 vs. < 60 years
old), type of fluoroquinolone, corticosteroid use, kidney dis-
Risk for tendon injuries
ease, and comparator (fluoroquinolone non-users vs. active
comparator). Risk window is the period of time considered
The treatment with fluoroquinolones was associated with an
after the end of fluoroquinolone treatment course during
increased risk of ATR (9 studies, OR 2.52 (95% CI 1.81–
which patients are at risk of suffering tendon injuries. The
3 . 5 2 ) , p < 0 . 0 0 1 , I 2 = 7 6 . 7 % ) ( F i g . 2 ; Ta b l e 2 ) .
influence of each individual study in overall risk estimate
Fluoroquinolones were also associated with an increased risk
was assessed using the “one-study removed” analysis. All
of AT (3 studies, OR 3.95 (95% CI 3.11–5.01), p < 0.001, I2 =
reported p values are 2-sided with significance being set as
0%) and of ATD (8 studies, OR 1.98 (95% CI 1.62–2.43),
less than 0.05. Stata (version 13.1) was used to perform
p < 0.001, I2 = 84.5%) (Fig. 3 and Fig. 4, respectively). The
statistics.
risk estimates did not significantly change when the analysis
was stratified according to different study designs, with the
exception of the risk of AT among cohort studies (1 study, OR
Results 3.70 (95% CI 0.90–15.16), p = 0.069). The increased risk of
all three outcomes remained statistically increased among
Figure 1 presents the flow of the search strategy criteria. older patients (aged ≥ 60 years) when the analysis was strati-
The electronic databases searched returned 1612 refer- fied by age. The risks estimated for any of the three outcomes
ences. After excluding duplicates and other studies with did not significantly change irrespective of the concomitant
inadequate design, 15 studies were included in this meta- use of corticosteroids and fluoroquinolones.
1434 Eur J Clin Pharmacol (2019) 75:1431–1443
Identification
1612 potentially relevant articles Relevant studies found by hand search
(PUBMED, n=426; EMBASE, n=1153; of reference lists
COCHRANE LIBRARY, n=33) (n = 0)
It was not possible to stratify the results according to No statistically significant result was identified among the
renal failure status, since only two studies assessed the studies evaluating the risk of ATR more than 30 days after the
risk of tendinopathy associated with fluoroquinolones in fluoroquinolone prescription (OR 1.72, 95% CI 0.70–4.25,
patients with renal insufficiency, one considering a differ- p = 0.238, I2 = 18.2%) (Table 2). The initial ATR risk estimate
ent outcome from the other (ATR vs. AT) [30, 35]. One did not significantly change among the studies considering a
identified an increased risk of AT (OR 20.0, 95% CI 2.7– 30-day at-risk window after the fluoroquinolone prescription.
149) and the other increased risk of ATR (RR 1.65, 95% When comparing fluoroquinolone users with non-users,
CI 1.32–1.84) [30, 35]. the initial increased risk estimates remained statistically sig-
The risk of AT did not significantly change when the nificant (Table 2). Only one study compared fluoroquinolones
analysis was conducted according to the methodological with an active comparator for both ATR and AT risk analyses
quality of the studies (Table 2). The risk of ATR remained (Table 2). The initial ATD risk estimate did not significantly
increased, with statistical significance, among the studies change when the analysis was stratified by different
with “moderate” methodological quality. Among the stud- comparators.
ies with “good” methodological quality, the risk of ATR The stratification of the risk estimates according to the
became statistically non-significant but associated with types of fluoroquinolones was only possible for the outcome
high heterogeneity (OR 1.94, 95% CI 0.76–4.99, p = ATR (Table 2). Ofloxacin and norfloxacin were associated
0.147, I2 = 91.0%). The risk of ATD became statistically with an increased risk of ATR, but not ciprofloxacin and
non-significant when only studies with “poor” methodo- levofloxacin. For AT and ATD, no sufficient data was avail-
logical quality were considered (OR 3.04, 95% CI 0.68– able to stratify the results according to the type of
13.52, p = 0.144, I2 = 72.0%). fluoroquinolone.
Table 1 Characteristics of the studies included
Intervention/cases Control
van der Linden et al. 1999 Retrospective cohort study Patients, aged at least 15 years old, living in The Netherlands, whose data Fluoroquinolone Reference group: amoxicillin,
was collected from 41 general practices in the period from January 1, trimethoprim, cotrimoxazole, or
1995 until December 31, 1996. All individuals treated with either nitrofurantoin
fluoroquinolones, amoxicillin, trimethoprim, cotrimoxazole, or
nitrofurantoin were followed from the first day of treatment until the
outcome of interest, death, transfer to another practice, or end of the
study period, whichever came first.
Eur J Clin Pharmacol (2019) 75:1431–1443
Sode et al. 2007 Retrospective cohort study Patients living at Funen County, Denmark, with data on Achilles tendon Fluoroquinolone Reference group: tetracyclines,
ruptures and fluoroquinolones, tetracyclines, penicillins, cephalosporins, penicillins, cephalosporins,
sulfonamides, macrolides, aminoglycosides, and glycopeptides use sulfonamides, macrolides,
retrieved from three population-based databases which included aminoglycosides, or
information on residents in primary and secondary care during the period glycopeptides
1991–1999.
Barge-Caballero et al. 2008 Retrospective cohort study Adult patients, aged at least 18 years old, submitted to heart transplantation Levofloxacin Levofloxacin non-use
in a Spanish hospital and prescribed with fluoroquinolones, with data
collected from August 1995 to September 2006
Hori et al. 2012 Retrospective cohort study Patients prescribed with fluoroquinolones or cephalosporins in Japan from Fluoroquinolone Cephalosporin
April 1996 to December 2009 were identified and considered for
inclusion in the study. Diagnosis of tendon disorder after
fluoroquinolone or cephalosporin administration was identified by an
automatic search of the hospital medical records.
Daneman et al. 2015 Retrospective cohort study The Ontario Registered Persons Database was used to identify adult Fluoroquinolone Amoxicillin
patients from Ontario, Canada, turning age 65 between April 1, 1997,
and March 31, 2012 were followed until primary outcome, death or end
of follow-up course.
Humbyrd et al. 2018 Retrospective cohort study Patients with end-stage renal disease (ESRD) using Medicare parts A and B Fluoroquinolone Fluoroquinolone non-use
(USA) as their primary health insurance between January 1, 1999, and
December 31, 2013. All patients were studied from their development of
ESRD until death, end of Medicare primary coverage, or end of the
study.
Jupiter et al. 2018 Retrospective cohort study Adult patients aged 18 to 90 years with at least one index prescription of Fluoroquinolone Fluoroquinolone non-use
fluoroquinolone between January 1, 2009, and June 30, 2014. Data was Reference group*
retrieved from Clinfomatics DataMart, USA. Propensity scores were
used to match fluoroquinolone users with other antibiotic users and
antibiotic non-users.
van der Linden et al. 2002 Case-control study Patients aged 18 years or over who had received a fluoroquinolone and had Diagnosis of ATR, AT, –
clinical data on IMS Health database, UK (July 1, 1992 to June 30, and ATD
1998). Potential cases were identified by reviewing patient profiles and
clinical data and excluded tendon disorders due to direct trauma. A group
of 10,000 control patients from the study cohort was randomly selected.
van der Linden et al. 2003 Case-control study Data was retrieved from the General Practice Research Database, UK, Diagnosis of ATR –
considering a period from 1988 to 1998. Cases were defined as all
persons who had an Achilles tendon rupture and who had at least
18 months of valid medical history before the index date. A group of
1435
Table 1 (continued)
1436
Study Risk period (days) Age mean (years) Female (%) Patients (n, ID/C) Outcomes/fluoroquinolone M.Q.
use (n, ID/C)*
van der Linden et al. 1999 30 days Fluo 53.0 Fluo 63.9 1841/9406 AT 4/4 7
Ref 45.0 Ref 71.4
Sode et al. 2007 90 days 44.2 25.0 28,262/401080 ATR 5/21 6
Barge-Caballero et al. 2008 30 days 58.8 22.1 149 ATD 14 3
Hori et al. 2012 30 days NR NR 17,147/38517 ATD 14/5 4
Daneman et al. 2015 30 days Fluo 65 Fluo 51.4 657,950/1,086,410 ATD 23028/14123 6
Amox 65 Amox 51.1
Humbyrd et al. 2018 NR NR 55.0 NR ATR: NR 5
Jupiter et al. 2018 30 days Fluo 43.5 Fluo 52.4 576,318/576,318/631304* ATD 550/323/551* 8
Non-U 43.8 Non-U 52.3
Ref 43.7 Ref NR
Eur J Clin Pharmacol (2019) 75:1431–1443
Eur J Clin Pharmacol (2019) 75:1431–1443 1437
ID, intervention drug; C, control; M.Q., methodological quality assessment; ATD, any tendon disorder; AT, Achilles tendinitis; ATR, Achilles tendon rupture; Fluo, fluoroquinolone; Ref, reference group;
Amox, amoxicillin; NR, not reported; ESRD, end-stage renal disease; USA, United States of America; UK, United Kingdom; THIN, The Health Improvement Network; CPRD, Clinical Practice Research
Among the results identified for the three outcomes, the
6
8
7
7
7
7
“one-study removed” analysis indicates that the overall risk
estimates did not significantly change when each study was
individually removed from the meta-analysis (Supplemental
Figs. 1, 2, and 3).
ATD 111/236
ATD 85/178
ATD 49/298
ATR 29/448
ATR 42/21
ATR 67/82
ATR 41/19
ATR 3/124
AT 267/77
AT 43/298
ATR 6/10
28,907/28,907
1367/50,000
4836/18356
3802/14002
947/18,940
742/10,000
Discussion
1118/1118
Controls 36.3
Controls 52.0
Controls: NR
Controls: NR
Cases 38.8
Cases 61.0
Cases 30.6
NR
Controls: NR
Controls: NR
Controls 47
Cases 55.9
Cases 56.0
Cases 61.6
Cases 46.5
NR
NR
30 days
30 days
30 days
30 days
.034 1 29.4
According to this meta-analysis, patients aged ≥ 60 years to the scarcity of data. Only two studies assessed the risk of
old are at an increased risk for ATR and AT. These results are tendinopathy associated with fluoroquinolones in patients suf-
in line with previous warnings issued by regulatory authori- fering from renal insufficiency, but considered two different
ties, which recommended that fluoroquinolones should be outcomes (ATR vs. AT), precluding a quantitative synthesis of
used with special caution in older patients [6, 8]. The influence the results [30, 35].
of concomitant use of corticosteroids on the risk of tendon Most of the studies assessing the risk of ATR due to
injury in patients treated with fluoroquinolones was also eval- fluoroquinolones included in this meta-analysis considered
uated in some pharmacoepidemiological studies. According risk-window periods of 30 days. Yet, two studies assumed
to the sensitivity analysis, it was observed an increase in risk risk-window periods of 90 days and two other studies did
sizes between 4.68-fold (ATD) and 14.72-fold (ATR) for the not report the length of time during which patients were con-
three outcomes among patients simultaneously using sidered to be at risk. The sensitivity analysis demonstrated that
fluoroquinolones and corticosteroids, compared with the ini- the risk of ATR remained statistically increased when the
tial estimates. Therefore, the combination of these drugs analysis was restricted to studies considering a 30-day risk
should be avoided, as previously recommended by drug reg- window, but not among those considering a risk period of
ulatory agencies [6, 8]. 90 days. A systematic review of case reports concluded that
The observational studies included in this meta-analysis 40% of the cases of tendon rupture were diagnosed within the
presented varying demographic characteristics and methodol- first 6 days after treatment initiation, while the mean time to
ogies. Two cohort studies investigating the risk of tendon in- onset was estimated at 26 days [2]. Additionally, half of the
juries associated with fluoroquinolones included heart cases of tendinitis occurred within the first 6 days of treatment
transplanted patients and end-stage renal disease patients and the mean time to onset of symptoms after the initiation of
[30, 32]. It has been suggested that individuals submitted to fluoroquinolone therapy was 18 days [2]. The studies included
organ transplant and patients with renal insufficiency are at in this meta-analysis assessed the risk of AT and ATD within a
greater risk of suffering tendon injuries, particularly Achilles risk period of 30 days.
tendon ruptures and tendinitis [30, 32]. Gender distribution of The control groups used by the observational studies
the population samples was significantly different among the included in this meta-analysis differed as well. Although
studies. There is conflicting evidence regarding the influence non-use of fluoroquinolones was the most common com-
of patients’ sex in the risk of tendon injuries. There are studies parator group, four cohort studies used active controls [15,
suggesting that one specific gender, male or female, may be a 27–29]. Noteworthy, it is possible that observational stud-
risk factor for tendon injury, while others have not [33–35]. ies using fluoroquinolone non-users as control groups may
Although renal impairment is a known risk factor for tendon have been affected by confounding-by-indication, namely
injury, it was not possible to conduct a sensitivity analysis due if the infections were causally related to the tendon injuries
Eur J Clin Pharmacol (2019) 75:1431–1443 1439
Outcome
Achilles tendon rupture 9 2.52 (1.81–3.52) < 0.001 76.7%
Achilles tendinitis 3 3.95 (3.11–5.01) < 0.001 0.0%
Any tendon disorders 8 1.98 (1.62–2.43) < 0.001 84.5%
Age
Achilles tendon rupture
< 60 years 3 1.71 (0.92–3.17) 0.088 16.0%
≥ 60 years 4 3.61 (1.21–10.79) 0.022 83.8%
Achilles tendinitis
< 60 years 2 1.35 (0.87–2.10) 0.180 31.8%
≥ 60 years 2 5.08 (1.93–13.33) 0.001 89.9%
Any tendon disorders
< 60 years 4 1.28 (1.11–1.49) 0.001 47.4%
≥ 60 years 4 2.17 (1.48–3.20) < 0.001 94.6%
Corticosteroid use
Achilles tendon rupture
Yes 5 14.72 (8.83–24.53) < 0.001 0.0%
No 4 2.27 (1.21–4.27) 0.011 59.2%
Achilles tendinitis
Yes 1 9.10 (4.60–18.00) < 0.001 –
No 1 3.20 (2.31–4.43) < 0.001 –
Any tendon disorders
Yes 4 4.68 (2.32–9.45) < 0.001 84.1%
No 2 1.56 (1.22–1.99) < 0.001 54.8%
Study designs
Achilles tendon rupture
Case controls 7 2.76 (1.86–4.08) < 0.001 71.3%
Cohorts 2 1.83 (1.04–3.24) 0.037 43.9%
Achilles tendinitis
Case controls 2 3.85 (2.75–5.37) < 0.001 41.6%
Cohorts 1 3.70 (0.90–15.16) 0.069 –
Any tendon disorders
Case controls 4 1.83 (1.46–2.30) < 0.001 65.6%
Cohorts 4 2.18 (1.57–3.02) < 0.001 87.4%
Methodological quality assessment
Achilles tendon rupture
Good methodological quality 2 1.94 (0.76–4.99) 0.167 91.0%
Moderate methodological quality 6 3.07 (2.36–3.98) < 0.001 0.0%
Poor methodological quality 1 1.56 (1.32–1.84) – –
Achilles tendinitis
Good methodological quality 0 – – –
Moderate methodological quality 3 3.95 (3.11–5.01) < 0.001 0.0%
Poor methodological quality 0 – – –
Any tendon disorder
Good methodological quality 2 1.68 (1.49–1.90) < 0.001 0.0%
Moderate methodological quality 4 2.08 (1.60–2.70) < 0.001 86.1%
Poor methodological quality 2 3.04 (0.68–13.52) 0.144 72.0%
1440 Eur J Clin Pharmacol (2019) 75:1431–1443
Table 2 (continued)
Studies Odds ratio (OR) Heterogeneity
n 95% CI p I2
Type of fluoroquinolone
Achilles tendon rupture
Ofloxacin 3 2.84 (1.31–6.19) 0.008 0.0%
Levofloxacin 1 0.60 (0.35–1.04) 0.069 –
Ciprofloxacin 3 1.34 (0.91–1.97) 0.138 0.0%
Norfloxacin 2 3.02 (1.32–6.93) 0.039 0.0%
Risk window
Achilles tendon rupture
≤ 30 days 5 3.32 (2.59–4.28) < 0.001 0.0%
> 30 days 2 1.72 (0.70–4.25) 0.238 66.7%
NR 2 1.68 (1.27–2.21) < 0.001 30.0%
Achilles tendinitis
≤ 30 days 3 3.95 (3.11–5.01) < 0.001 0.0%
> 30 days 0 – – –
NR 0 – – –
Any tendon disorder
≤ 30 days 8 1.98 (1.62–2.43) < 0.001 84.5%
> 30 days 0 – – –
NR 0 – – –
Comparator
Achilles tendon rupture
Fluoroquinolone non-users 8 1.58 (1.27–1.96) < 0.001 0.0%
Active comparator 1 3.10 (0.13–72.34) 0.481 –
Achilles tendinitis
Fluoroquinolone non-users 2 3.77 (1.48–9.57) 0.005 0.0%
Active comparator 1 3.70 (0.003–483.70) 0.718 –
Any tendon disorder*
Fluoroquinolone non-users 7 1.88 (1.57–2.24) < 0.001 41.4%
Active comparator 2 1.18 (1.03–1.35) 0.016 0.0%
Data source
Achilles tendon rupture# 7 2.28 (1.57–3.30) < 0.001 74.2%
Achilles tendinitis 3 3.95 (3.11–5.01) < 0.001 0.0%
Any tendon disorders$ 7 2.06 (1.65–2.56) < 0.001 85.2%
*One study (Jupiter et al. 2018) assessed the risk of ATD associated with fluoroquinolones vs. non-users and vs.
active comparator
#
van der Linden et al. (2003) (MQ score 7) and Persson and Jick (MQ score 7) were removed from the analysis
because their data source overlapped with van der Linden et al. (2002) (MQ score 7) and Morales et al. (2018)
(MQ score 8), respectively (Note that although van der Linden et al. (2002) and van der Linden et al. (2003) both
have the same MQ score, the 2003 publication was removed because it is the most recent one)
$
Persson and Jick was removed from the analysis because the data source overlapped with Morales et al. (2018)
events [15]. Therefore, some authors established compari- studies used antibiotics as negative controls in order to
sons with active controls. Yet, the studies included in this assess the specificity of the association between
meta-analysis used different antibiotics as active controls. fluoroquinolones and tendinopathy [9, 15]. Nevertheless,
Some compared fluoroquinolone users with a composite of the risk of tendon injuries seems to increase due to
multiple antibiotics, such as nitrofurantoin, amoxicillin, fluoroquinolones use, regardless of the chosen comparator.
tetracyclines, and trimethoprim, which are mostly indicat- Only three out of the 15 studies were assessed as having
ed to treat mild urinary tract infections [15, 27, 28]. Two “good” methodological quality. Most of them did not take into
Eur J Clin Pharmacol (2019) 75:1431–1443 1441
.066 1 15.2
account all the factors known to modify the risk of tendon studies that were also included in this systematic review and
injuries, such as trauma, transplant, obesity, diabetes mellitus, meta-analysis [10, 12].
kidney failure, or recreational sports. The reduced risk of con- The systematic review and meta-analysis conducted by
founding was deemed to be important in order to attain higher Yu and colleagues (2019) have identified an increased risk
methodological scores. These may have been the reason why of tendinopathy associated with fluoroquinolones as well
previous systematic reviews using the same methodological [12]. However, some methodological approaches which
quality assessment tool awarded higher scores to some of the could strengthen the analysis were not considered by the
.0573 1 17.4
11. van der Vlist AC, Breda SJ, Oei EHG, Verhaar JAN, de Vos RJ 28. Sode J, Obel N, Hallas J, Lassen A (2007) Use of fluroquinolone
(2019) Br J Sports Med. https://doi.org/10.1136/bjsports-2018- and risk of Achilles tendon rupture: a population-based cohort
099991 study. Eur J Clin Pharmacol 63:499–503
12. Yu X, Jiang DS, Wang J, Wang R, Chen T, Wang K, Cao S, Wei X 29. Hori K, Yamakawa K, Yoshida N, Ohnishi K, Kawakami J (2012)
(2019) Fluoroquinolone use and the risk of collagen-associated ad- Detection of fluoroquinolone-induced tendon disorders using a hos-
verse events: a systematic review and meta-analysis. Drug Saf. pital database in Japan. Pharmacoepidemiol Drug Saf 21:886–889
https://doi.org/10.1007/s40264-019-00828-z 30. Humbyrd CJ, Bae S, Kucirka LM, Segev DL (2018) Incidence, risk
13. Persson R, Jick S (2019) Clinical implications of the association factors, and treatment of Achilles tendon rupture in patients with
between fluoroquinolones and tendon rupture: the magnitude of the end-stage renal disease. Foot Ankle Int 39:821–828
effect with and without corticosteroids. Br J Clin Pharmacol 85: 31. Jupiter DC, Fang X, Ashmore Z, Shibuya N, Mehta HB (2018) The
949–959 relative risk of Achilles tendon injury in patients taking quinolones.
14. Nyyssönen T, Lantto I, Lüthje P, Selander T, Kröger H (2018) Drug Pharmacotherapy 38:878–887
treatments associated with Achilles tendon rupture. A case-control 32. Barge-Caballero E, Crespo-Leiro MG, Paniagua-Martín MJ, Muñiz
study involving 1118 Achilles tendon ruptures. Scand J Med Sci J, Naya C, Bouzas-Mosquera A, Piñón-Esteban P, Marzoa-Rivas R,
Sports 28:2625–2629 Pazos-López P, Cursack GC, Cuenca-Castillo JJ, Castro-Beiras A
15. Daneman N, Lu H, Redelmeier DA (2015) Fluoroquinolones and (2008) Quinolone-related Achilles tendinopathy in heart transplant
collagen associated severe adverse events: a longitudinal cohort patients: incidence and risk factors. J Heart Lung Transplant 27:46–
study. BMJ Open 5:e010077 51
16. Ioannidis JP, Patsopoulos NA, Rothstein HR (2008) Reasons or 33. van der Linden PD, Sturkenboom MC, Herings RM, Leufkens HM,
excuses for avoiding meta-analysis in forest plots. BMJ 336: Rowlands S, Stricker BH (2003) Increased risk of Achilles tendon
1413–1415 rupture with quinolone antibacterial use, especially in elderly pa-
17. Smith TC, Spiegelhalter DJ, Thomas A (1995) Bayesian ap- tients taking oral corticosteroids. Arch Intern Med 163:1801–1807
proaches to random-effects meta-analysis: a comparative study. 34. Seeger JD, West WA, Fife D, Noel GJ, Johnson LN, Walker AM
Stat Med 14:2685–2699 (2006) Achilles tendon rupture and its association with fluoroquin-
18. Berlin JA (1995) Invited commentary: benefits of heterogeneity in olone antibiotics and other potential risk factors in a managed care
meta-analysis of data from epidemiologic studies. Am J Epidemiol population. Pharmacoepidemiol Drug Saf 15:784–792
142:383–387 35. Wise BL, Peloquin C, Choi H, Lane NE, Zhang Y (2012) Impact of
19. Moher D, Liberati A, Tetzlaff J, Altman DG, PRISMA Group age, sex, obesity, and steroid use on quinolone-associated tendon
(2009) Preferred reporting items for systematic reviews and meta- disorders. Am J Med 125:1228.e23–1228.e28
analyses: the PRISMA statement. PLoS Med 6:e1000097
36. Corrao G, Zambon A, Bertù L et al (2006) Evidence of tendinitis
20. U.S. National Library of Medicine. Medical Subject Headings
provoked by fluoroquinolone treatment: a case-control study. Drug
2019. U.S. National Library of Medicine. https://meshb.nlm.nih.
Saf 29:889–896
gov/search. Accessed 14 May 2019
37. van der Linden PD, Sturkenboom MC, Herings RM, Leufkens HG,
21. ICH MedDRA Maintenance and Support Services Organization
Stricker BH (2002) Fluoroquinolones and risk of Achilles tendon
(MSSO). International Council for Harmonisation of technical re-
disorders: case-control study. BMJ 324:1306–1307
quirements for Pharmaceuticals for Human use (ICH). https://www.
38. Williams RJ 3rd, Attia E, Wickiewicz TL, Hannafin JA (2000) The
meddra.org/. Accessed 14 May 2019
effect of ciprofloxacin on tendon, paratenon, and capsular fibroblast
22. Elsevier. Embase Indexing and Emtree. https://www.elsevier.com/
metabolism. Am J Sports Med 28:364–369
solutions/embase-biomedical-research. Accessed 14 May 2019
23. Wells GA, Shea B, O’Connell D et al (2014). The Newcastle- 39. Tsai WC, Hsu CC, Chen CP et al (2011) Ciprofloxacin up-regulates
Ottawa Scale (NOS) for assessing the quality of nonrandomised tendon cells to express matrix metalloproteinase-2 with degradation
studies in meta-analyses. http://www.ohri.ca/programs/clinical_ of type I collagen. J Orthop Res 29:67–73
epidemiology/oxford.asp. Accessed 14 May 2019 40. Shakibaei M, de Souza P, van Sickle D, Stahlmann R (2001)
24. DerSimonian R, Laird N (1986) Meta-analysis in clinical trials. Biochemical changes in Achilles tendon from juvenile dogs after
Control Clin Trials 7:177–188 treatment with ciprofloxacin or feeding a magnesium-deficient diet.
25. Higgins JP, Thompson SG, Deeks JJ, Altman DG (2003) Arch Toxicol 75:369–374
Measuring inconsistency in meta-analyses. BMJ 327:557–560 41. Pouzaud F, Bernard-Beaubois K, Thevenin M, Warnet JM, Hayem
26. B Publication bias (2009) In: Borenstein M, Hedges LV, JPT H, G, Rat P (2004) In vitro discrimination of fluoroquinolones toxicity
Rothstein HR (eds) Introduction to meta-analysis. Wiley, on tendon cells: involvement of oxidative stress. J Pharmacol Exp
Chichester, pp 277–291 Ther 308:394–402
27. van der Linden PD, van de Lei J, Nab HW, Knol A, Stricker BH
(1999) Achilles tendinitis associated with fluoroquinolones. Br J Publisher’s note Springer Nature remains neutral with regard to
Clin Pharmacol 48:433–437 jurisdictional claims in published maps and institutional affiliations.